\begin{table}
\centering
\footnotesize
\caption{Comparison results, British election study} \vspace{0.5cm}
\begin{tabular}{l|cc|cc|cc|cc|c}
 & STR & STR & STR & STR & & & BES & BES & \\
\textbf{Predictor} & \textbf{Empty} & \textbf{Naive} & \textbf{Naive, rn} & \textbf{Within} & $\% 1$ & $\% 2$ & \textbf{Empty} & \textbf{Naive} & $\Delta$ \\ \hline
Education years &        0.126 &        0.103 &        0.103 &        0.034 &         81.6\% &         32.6\% &        0.143 &        0.119 &         83.4\% \\
 & (        0.009 ) & (        0.014 ) & (        0.014 ) & (        0.021 ) & (t=-1.408) & (t=-2.774) & (        0.003 ) & (        0.004 ) & (t=-4.488) \\
College &        0.120 &        0.086 &        0.086 &        0.052 &         71.5\% &         60.2\% &        0.132 &        0.130 &         97.8\% \\
 & (        0.010 ) & (        0.014 ) & (        0.014 ) & (        0.019 ) & (t=-2.009) & (t=-1.437) & (        0.006 ) & (        0.006 ) & (t=-.346) \\
Work income &        0.093 &        0.044 &        0.061 &        0.019 &         47.7\% &         30.5\% &        0.117 &        0.112 &         95.9\% \\
 & (        0.010 ) & (        0.013 ) & (        0.013 ) & (        0.016 ) & (t=-2.879) & (t=-2.094) & (        0.004 ) & (        0.005 ) & (t=-.759) \\
\hline
\end{tabular}

\vspace{.25cm}

Averages of beta coefficients for 8 shared political issue variables (standardized by positive sign in the empty model).
\end{table}
